Regulation of Hate Speech and Pornography After RAV PDF

Regulation of Hate Speech and Pornography after R.A.V.
Author(s): Elena Kagan Source: The University of Chicago Law Review, Vol. 60, No. 3/4 (Summer - Autumn, 1993), pp. 873-902 Published by: The University of Chicago Law Review Stable URL: http://www.jstor.org/stable/1600159 Accessed: 24/02/2010 17:21
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=uclr. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
The University of Chicago Law Review is collaborating with JSTOR to digitize, preserve and extend access to The University of Chicago Law Review.
http://www.jstor.org
Regulation of Hate Speech and Pornography After R.A.V.

Elena Kagant This Essay on the regulation of hate speech and pornography addresses both practicalities and principles. I take it as a given that we live in a society marred by racial and gender inequality, that certain forms of speech perpetuate and promote this inequality, and that the uncoerced disappearance of such speech would be cause for great elation. I do not take it as a given that all governmental efforts to regulate such speech thus accord with the Constitution. What is more (and perhaps what is more important), the Supreme Court does not, and will not in the foreseeable future, take this latter proposition as a given either. If confirmationof this point were needed, it came last year in the shape of the Court's opinion in R.A.V. v City of St. Paul.' There, the Court struck down a so-called hate speech ordinance, in the process reiterating, in yet strengthened form, the tenet that the First Amendment presumptively prohibits the regulation of speech based upon its content, and especially upon its viewpoint. That decision demands a change in the nature of the debate on pornography and hate speech regulation. It does so for principled reasons-because it raises important and valid questions about which approaches to the regulation of hate speech and pornography properly should succeed in the courts. And it does so for purely pragmatic reasons-because it makes clear that certain approaches almost surely will not succeed. In making this claim, I do not mean to suggest that all efforts to regulate pornography and hate speech be suspended, on the
t Assistant Professor of Law, The University of Chicago Law School. This Essay is based on remarks I made in a panel discussion at the conference,"Speech, Equality, and Harm: Feminist Legal Perspectives on Pornographyand Hate Propaganda,"held at The University of Chicago Law School on March 5-7, 1993. The argument has been expanded only slightly, and the reader is asked to make allowances for my necessarily abbreviated discussion of many complicatedissues. I am gratefulto Mary Becker, LarryLessig, Michael McConnell,GeoffreyStone, David Strauss, and Cass Sunstein for valuable advice and comments. The Class of '64 Fund and the Russell Parsons Faculty ResearchFund at The University of Chicago Law School provided financial support. 1 112 S Ct 2538 (1992).
873
874
The University of Chicago Law Review
[60:873
ground either of mistake or of futility. Quite the opposite. R.A.V. largely forecloses some lines of advocacy and argument (until now the dominant lines), as well perhaps it should have. But the decision leaves open alternative means of regulatingsome pornography and hate speech, or of alleviating the harms that such speech causes. The primarypurpose of this Essay is to offer some of these potential new approachesfor considerationand debate. The question I pose is whether there are ways to achieve at least some of the goals of the anti-pornography and anti-hate speech movements without encroachingon valuable and ever more firmly settled First Amendment principles. This Essay is just that-an essay, a series of trial balloons, which may be shot down, from either side or no side at all, by me or by others. The point throughoutis to emphasize the range of approachesremaining available after R.A.V. and meriting discussion.
I. THE PROBLEM OFVIEWPOINT DISCRIMINATION
In R.A.V., the Court struck down a local ordinance construed to prohibit those fighting words, but only those fighting words, based on race, color, creed, religion, or gender.2 Fighting words long have been considered unprotected expression-so valueless and so harmful that governmentmay prohibit them entirely without abridging the First Amendment.3Why, then, was the ordinance before the Court constitutionally invalid? The majorityreasoned that the ordinance's fatal flaw lay in its incorporationof a kind of content-based distinction. The ordinance,on its very face, distinguished among fighting words on the basis of their subject matter: only fighting words concerning"race, color, creed, religion or gender" were forbidden.4More, and much more nefariously in the Court's view, the ordinance in practice discriminatedbetween different viewpoints: it effectively prohibited racist and sexist fighting words, while allowing all others.5Antipathy to such viewpoint distinctions, the Court stated, lies at the heart of the guarantee of freedom of expression. "The government may not regulate [speech] based on hostility-or favoritism-towards the underlying
2 Id at 2542. The Supreme Court defined "fighting words" in Chaplinsky v New Hampshire, 315 US 568, 572 (1942), as words "which by their very utterance inflict injury or tend to incite an immediate breach of the peace." 3 Chaplinsky, 315 US at 572. 4 R.A.V., 112 S Ct at 2541, 2547. 6 Id at 2547-48.
1993]
RegulationAfter R.A.V.
875
message expressed"; it may not suppress or handicap "particular ideas."6 The reasoning in R.A.V. closely resembles that found in the key judicial decision on the regulation of pornography.In American Booksellers Ass'n, Inc. v Hudnut,7 affirmed summarily by the Supreme Court, the United States Court of Appeals for the Seventh Circuit invalidated the Indianapolis anti-pornography ordinance drafted by Andrea Dworkin and Catharine MacKinnon. That ordinance declared pornographya form of sex discrimination, with pornographydefined as "the graphic sexually explicit subordination of women, whether in pictures or in words," that depicted women in specified sexually subservient postures.8The core problem for the Seventh Circuit, as for the Supreme Court in R.A.V., was one of viewpoint discrimination. The ordinance, according to the Court of Appeals, made the legality of expression "depend[ent] on the perspective the author adopts."9 Sexually explicit speech portraying women as equal was lawful; sexually explicit speech portrayingwomen as subordinate was not. The ordinance, in other words, "establishe[d] an 'approved' view" of women and of sexual relations.10 From this feature, invalidation necessarily followed: "The state may not ordain preferred viewpoints in this way. The Constitution forbids the state to declare one perspective right and silence opponents."1' The approach used in R.A.V. and Hudnut has a large body of case law behind it. The presumption against viewpoint discrimination did not emerge alongside of, or in response to, the effort to curtail certain forms of racist and sexist expression. Rather, that presumption long has occupied a central position in First Amendment doctrine. Decades ago, for example, the Supreme Court employed the presumption to strike down laws restricting expression that discredited the military or that presented adultery in a favorable light, and more recently, the Court invoked the presumption to invalidate flag-burning statutes.l2 This is not to say that the Court invariably has invalidated laws that incorporate view6 Id at 2545, 2549.
7 8
771 F2d 323 (7th Cir 1985), affd mem, 475 US 1001 (1986).
Id at 324.
9 Id at '0 Id. l2
328.
n Id at 325.
See Schacht v United States, 398 US 58, 67 (1970) (military);Kingsley Int'l Pictures Corp. v Regents, 360 US 684, 688 (1959) (adultery);Texas v Johnson, 491 US 397, 416-17 (1989) (flag-burning);United States v Eichman, 496 US 310, 317-18 (1990) (same).
876
The University of ChicagoLaw Review
[60:873
point favoritism. Exceptions to the rule exist, although the Court rarely has seen fit to acknowledgethem as such; in a number of areas of First Amendment law (and especially when so-called lowvalue speech is implicated), the Court breezily has ignored both more and less obvious forms of viewpoint preference.13 Still, the rule has been more often honored than honored in the breach,and the Supreme Court's opinion in R.A.V., as well as its summaryaffirmance of Hudnut, could have been expected. Moreover,the Court's decision in R.A.V. entrenched still further the presumptionagainst viewpoint-basedregulationof speech. To be sure, the majorityopinion received only five votes and came under vehement attack from the remainingJustices.14Thus, some might reason that the disposition of the case reveals a weakening in the Court's commitment to viewpoint neutrality, either across the board or with respect to racist and sexist expression. If this reasoningwere valid, those disliking R.A.V. might simply wait and pray for an advantageouschange in the Court's membership.But any such reading of the case rests on a grave misunderstanding. The Court's opinion received the support of only a bare majority because, for two reasons having nothing to do with the particular viewpoint involved, the case appeared to some Justices not to invoke the presumption against viewpoint regulation at all. First,
13 Several examples of this blindness to viewpoint discrimination occur in the area of commercial speech. See Posadas de Puerto Rico Associates v Tourism Co. of Puerto Rico, 478 US 328, 330-31 (1986) (upholding a law prohibiting advertising of casino gambling, but leaving untouched all speech discouraging such gambling); Central Hudson Gas & Electric v Public Service Commission, 447 US 557, 569-71 (1980) (striking down a broad law prohibiting advertising to stimulate the use of electricity, but suggesting that a more narrowly-tailored law along the same lines would meet constitutional standards, even if the law were to allow all expression discouraging use of electricity). In addition, as Catharine MacKinnon has noted, the delineation of entire low-value categories of speech, such as obscenity and child pornography, may be thought to reflect a kind of viewpoint discrimination, given that the speech falling within such categories likely expresses a single (disfavored) viewpoint about sexual matters. See Catharine A. MacKinnon, Feminism Unmodified: Discourses on Life and Law 212 (Harvard, 1987). Further discussion of this point, and its relevance for the regulation of pornography and hate speech, appears in note 73 and the text accompanying note 80. Finally, the Court has indicated that the usual presumption against viewpoint discrimination does not apply, or at least does not apply in full force, when the government engages in selective funding of speech, rather than selective restriction of speech. See Rust v Sullivan, 111 S Ct 1759, 1772-73 (1991); text accompanying notes 28-29. '1 The four Justices who refused to join the Court's opinion also voted to invalidate the St. Paul ordinance, but only because of a concern about overbreadth that easily could have been corrected. They assailed the majority's conclusion that the presumption against viewpoint discrimination mandated invalidation of the statute, either on the view that the presumption failed to operate in spheres of unprotected speech, see 112 S Ct at 2551-54 (White concurring) and id at 2560 (Blackmun concurring), or on the view that the ordinance incorporated no viewpoint-based distinction, see id at 2570-71 (Stevens concurring).
1993]
Regulation After R.A.V.
877
and most important, the alleged viewpoint discrimination in the case occurred within a category of speech-fighting words-that the Court long ago declared constitutionally unprotected. Second, the viewpoint discrimination found in the ordinance existed not on its face, but only in application-and even in application, only with a fair bit of argument.15 Had the law distinguished on its face between racist (or sexist) speech and other speech outside the category of fighting words, the Court's decision likely would have been unanimous.l6What R.A.V. shows, then, is the depth, not the tenuousness, of the Court's commitment to a viewpoint neutrality principle. And what R.A.V. did, in applying that principle to a case of non-facial discrimination in an unprotected sphere, was to render that principle even stronger. Any attempt to regulate pornography or hate speech-or at least any attempt standing a chance of success-must take into account these facts (the "is," regardless whether the "ought") of First Amendment doctrine. A law specifically disfavoring racist or sexist speech (or, to use another construction, a law distinguishing between depictions of group members as equal and depictions of group members as subordinate) runs headlong into the longstanding, and newly revivified, principle of viewpoint neutrality. I do not claim that exceptions to this principle will never be made, or even that such exceptions will not be made by the current Court. Exceptions, as noted previously, have been recognized before (even if not explicitly); they doubtless will be recognized again; and in the last section of this Essay, I consider briefly whether and how to frame them. I do claim that given the current strength of the viewpoint neutrality principle, a purely pragmatic approach to regulating hate speech and pornographywould seek to use laws not subject to the viewpoint discrimination objection, while also seeking to justify-as exceptions-carefully crafted and limited departuresfrom the rule against viewpoint regulation.
16 The St. Paul ordinance,on its face, discriminatedonly on the basis of subject matter, as the Court conceded. For the dispute on whether the ordinanceapplied in a viewpointdiscriminatorymanner,contrast the majorityopinion, 112 S Ct at 2547-48,with the concurring opinion of Justice Stevens, id at 2570-71.Contrastalso Cass R. Sunstein, On Analogical Reasoning, 106 Harv L Rev 741, 762-63 & n 78 (1993) (R.A.V. ordinance not viewpoint-
based in practice), with Elena Kagan, The Changing Faces of First Amendment Neutrality: R.A.V. v St. Paul, Rust v Sullivan, and the Problem of Content-Based Underinclusion, 1992
S Ct Rev 29, 69-71 (R.A.V. ordinanceviewpoint-basedin practice). 16 See note 14 for a description of the concurringJustices' objections to the Court's decision. In the case hypothesized in the text, those objections would have evaporated.
878
[60:873
This approach, in my view, also best accords with important free speech principles (the "ought"in the "is" of First Amendment doctrine). A focus on the feasible is arguably irresponsibleif the feasible falls desperately short of the proper. But here, I think, that is not the case. If reality-the current state of First Amendment doctrine-counsels certain proposals and not others, certain lines of argument and not others, so too do important values embodied in that doctrine. More specifically, the principle of viewpoint neutrality, which now stands as the primary barrierto certain modes of regulating pornographyand hate speech, has at its core much good sense and reason. Although here I can do no more than touch on the issue, my view is that efforts to regulate pornography and hate speech not only will fail, but also should fail to the extent that they trivialize or subvert this principle. Those who have criticized the courts for using the viewpoint neutrality principle against efforts to regulate pornographyor hate speech usually have offered one of two arguments.First, some have claimed that such efforts comportwith the norm of viewpoint neutrality because they are based on the harm the speech causes, ratherthan the viewpoint it espouses.17 Second, and more dramatically, some have challenged the norm itself as incoherent, worthless, or dangerous.l8Both lines of argument have enriched discussion of the viewpoint neutrality principle, by challenging the tendency of such discussion to do nothing more than apotheosize. Yet both approaches, in somewhat different ways, slight the reasons and values underlying current First Amendment doctrine-including the decisions in R.A.V. and Hudnut. The claim that pornography and hate-speech regulation is harm-based, rather than viewpoint-based, has an initial appeal, but turns out to raise many hard questions. The claim appeals precisely because it reflects an understandingof the value of a view-
See, for example, Cass R. Sunstein, Pornography and the First Amendment, 1986 Duke L J 589, 612; MacKinnon, Feminism Unmodified at 212 (cited in note 13). See also R.A.V., 112 S Ct at 2570 (Stevens concurring). Professor Sunstein always has combined this argument with a fuller analysis of when exceptions to the viewpoint regulation doctrine are justified; for him, the ability to classify a law as harm-based seems not the end, but only the start of the inquiry. See Cass R. Sunstein, Words, Conduct, Caste, 60 U Chi L Rev 795, 796 (1993) (in this issue). My brief discussion, in Section II of this Essay, on whether and when to recognize such exceptions owes much to his work on the subject. 18 See Mary Becker, Conservative Free Speech and the Uneasy Case for Judicial Review, 64 U Colo L Rev 975, 1044-47 (1993) (arguing that a viewpoint neutrality norm harms women and minority groups); MacKinnon, Feminism Unmodified at 210-13 (cited in note 13) (challenging the ability to identify viewpoint regulation except by reference to social consensus).
17
1993]
879
point neutrality norm and a desire to maintain it: if pornography and hate-speech regulation is harm-based, then we can have both But the two yearnit and a rule against viewpoint discrimination.19 ings may not be so easy to accommodate, for it is not clear that the classification proposed can support much weight. It is true that statutory language can focus either on the viewpoint of speech or on the injury it causes: contrast an ordinance that prohibits "sexually explicit materials approving the subordination of women" with an ordinance that prohibits "sexually explicit materials causing the But if we assume (as a meaningful syssubordination of women."20 tem of free speech must) that speech has effects-that the expression of a view will often cause people to act on it-then the two phrasings should be considered identical for First Amendment purposes. To grasp this point, consider here a few further examples. Contrast a law that prohibits criticism of the draft with a law that prohibits any speech that might cause persons to resist the draft.2' Or, to use a case with more contemporaryresonance, contrast an ordinance punishing abortion advocacy and counseling with an ordinance punishing any speech that might induce a woman to get an abortion. To sever these pairs of statutes would be to transform the First Amendment into a formal rule of legislative drafting, concerned only with appearance. In all these cases, the facially harm-based statute and the facially viewpoint-based statute function in the same way, because it is speech of a certain viewpoint, and only of that viewpoint, which causes the alleged injury. The facially harm-based statute in these circumstances will curtail expression of a particular message as surely as will the statute that refers to the message in explicit language. Given this functional identity, the statutes properly are viewed as cognates.22
19 I suspect that a wish of this kind explains Justice Stevens's insistence in R.A.V. that the St. Paul ordinanceregulated speech "not on the basis of ... the viewpoint expressed, but rather on the basis of the harm the speech causes." 112 S Ct at 2570 (Stevens concurring). Both in R.A.V. and in numerousother opinions and articles, Justice Stevens has expressed unwaveringsupport for the presumptionagainst viewpoint regulation.For the most recent example, see The Hon. John Paul Stevens, The Freedom of Speech, 102 Yale L J 1293, 1309 (1993). 20 The example, in slightly different form, appears in GeoffreyR. Stone, Anti-Pornog9 Harv J L & Pub Pol 461, 467 (1986). As raphy Legislation as Viewpoint-Discrimination, Stone points out, the MacKinnon-Dworkin ordinance,as written, is at any rate closer to the law focusing on the viewpoint espoused than to the law focusing on the harm caused. Id. 21 This example also appears in Stone. Id. 22 An argumentto the contrarymight rely not on the effects of the statutes, but on the intent of the legislature in passing them. The claim here would be that the facially harmbased statute more likely springs from a legitimate governmental motive than does the facially viewpoint-based statute. But this claim seems dubious in any case in which the
880
[60:873
This equivalencedoes not by itself destroy the claim that pornography regulation is harm-based, because both versions of the law might be characterizedin this manner:so long as a legislature reasonablydecides, as it surely could with respect to pornography, that speech causes harm, then regulation respondingto that harm (howeverframed) might be considered neutral, rather than an effort to disfavor certain viewpoints. But this approach,too, makes any distinction between viewpoint-based regulation and harmbased regulationcollapse upon itself. Using this analysis, almost all viewpoint-based regulation can be described as harm-based, responding neutrally not to ideas as such, but to their practical consequences. For it is difficult to see why anyone would opt to regulate a viewpoint that did not cause what seemed (to the regulators at least) to be a harm-or at a bare minimum, that could not reasonably be described as harmful. So, to return to the examples used above, a law prohibiting criticism of the draft could be termed harm-basedgiven that such speech in fact produces draft resistance; or a law prohibiting abortion counseling and advocacy could be termed harm-based given that such speech in fact increases the incidence of abortion (which many would count a serious injury). The substitution of labels-"harm-based" for "viewpoint-based"-thus either allows most viewpoint regulation to go forward or leaves yet unanswered the central issue of precisely when such regulation is appropriate. The more extreme critique of a case like Hudnut-that viewpoint discriminationdoctrine is both incoherentand corrupt-is in many ways more difficult to counter. This critique rebels against the very core of First Amendment doctrine by accepting the government's power to suppress viewpoints as such whenever the viewpoints are thought to cause some requisite harm.23But the justificationfor this position includes at least one extremely potent point: that recognizing viewpoint regulation may well depend on the decisionmaker's viewpoint; more specifically, that a judicial
statutes in fact operate in a similar manner. Because the legislators will know that the facially harm-based statute, like the facially viewpoint-based statute, will succeed in curtailing a specific message, their decision to phrase the statute in terms of harm (especially in light of a legal rule that effectively counsels them to do so) cannot provide a guarantee of legitimate intent. 23 See MacKinnon, Feminism Unmodified at 212-13 (cited in note 13). Even under current First Amendment doctrine, the government may engage in viewpoint discrimination in emergency circumstances amounting to something like a clear and present danger. The critique discussed in the text would allow viewpoint regulation on a much less stringent showing.
1993]
881
decisionmaker will be least likely to recognize (or count as relevant) viewpoint regulation when the regulator'sviewpoint lines up with his own.24This phenomenon may explain in part the willingness of courts to accept anti-obscenity laws at the same time as they strike down anti-pornography laws.25 More generally, this epistemological problem may skew viewpoint discrimination doctrine, as it operates in practice, in favor of the status quo-resulting in the disproportionate approval of laws most reflective of traditional sentiment and the disproportionateinvalidation of laws least so. But even assuming this is true, I doubt that the appropriate response lies in undermining, let alone eliminating, the viewpoint discrimination principle. That principle grows out of two concerns, as meaningful today as ever in the past.26The first relates to the effects of viewpoint discrimination: such action skews public debate on an issue by restricting the ability of one side (and one side only) to communicate a message. The second relates to governmental purposes: viewpoint regulation often arises from hostility toward ideas as such, and this disapproval constitutes an illegitimate justification for governmental action. Of course, particular instances of viewpoint discrimination may spring from benign purposes and have benign effects. Legislators may engage in viewpoint discrimination in an effort not to suppress ideas, but to respond to real harms; and the resulting damage to public discourse may signify little when measured against the harms averted. But how are the courts, or the people, or even legislators themselves to make these determinations of motive and effect in any given case? Will it not always be true that a benign motive can be assigned to governmental action? Will not any judgment as to relative harms depend on an evaluation of the message affected? From these questions, relating to the difficulty of evaluating particular purposes and effects, emerges a kind of rule-utilitarianjustification for the ban on viewpoint discrimination. The historic examples of the dangers of viewpoint discrimination, on the counts of both purpose and effect, are well-known and legion: the government's attempts, especially during World War I, to stifle criticism of military activities; its attempts in the 1950s to
See id at 212; Becker, 64 U Colo L Rev at 1046-47 (cited in note 18). 25 For discussion of the viewpoint bias inherent in obscenity laws, see notes 13 and 73 and text accompanyingnote 80. 26 The classic discussion of the bases for viewpoint discriminationdoctrine is Geoffrey
24
R. Stone, Content Regulation and the First Amendment, 25 Wm & Mary L Rev 189 (1983).
882
[60:873
suppress support of Communism;its efforts, stretching over decades, to prevent the burning of Americanflags as a means of proAnd if all these seem retesting the governmentand its policies.27 mote either from current threats or from the kind of viewpoint regulationat issue in Hudnut and R.A.V.-if they seem the stories of another generation,with little relevance for today-consider instead the case of Rust v Sullivan,28previewedin earlier hypotheticals. There, the government favored anti-abortion speech over abortion advocacy, counseling, and referral, and the Court, to its discredit, announced that because the selectivity occurred in the context of a governmental funding program, the presumption Or instead conagainst viewpoint discriminationwas suspended.29 sider the numerous ways in which some of the strange bedfellows of anti-pornography feminists (and one must admit their presence) might choose (indeed, have chosen) to attack the expression of, among others, gays and lesbians. The key point here is only strengthened by the insight that viewpoint discriminationdoctrine, as applied by the courts, has a way of producingsome patterned inconsistencies;or to put this another way, the very critique of the Court's viewpoint discrimination doctrine exposes the need for a viewpoint neutrality principle. For what the critique highlights is the tendency of governmental actors (of all kinds) to see speech regulation through the lens of their own orthodoxies, as well as the ease with which such orthodoxies can thereby become entrenched. Recognition of this process lies at the very core of the viewpoint discriminationdoctrine: as Justice Stevens recently has noted, that doctrine responds, preeminently, to fear of the "imposition of an official oreven (or perhaps especially) as to matters involvingsex thodoxy,"30 or race. That judicial decisionmakers, in applying the doctrine, sometimes will succumb to the views they hold hardly argues in favor of grantingcarte blanche to legislative decisionmakersto bow to theirs. It is difficult to see how women and minorities,who have the most to lose from the establishment of political orthodoxy,
27
See Akhil Reed Amar, The Supreme Court, 1991 Term-Comment:
The Case of the
R.A.V.v. City of St. Paul, 106 HarvL Rev 124 (1992),for a compariMissing Amendments: son of R.A.V.and the Court'smost recent flag-burning cases, Texas v Johnson, 491 US 397
(1989), and United States v Eichman, 496 US 310 (1990). 28 111 S Ct 1759 (1991).
29 Id at 1771-73.For a comparisonof Rust and R.A.V., see Kagan, 1992 S Ct Rev 29 (cited in note 15). 30 Stevens, 102 Yale L J at 1304 (cited in note 19).
1993]
883
would gain by jettisoning the First Amendment doctrine that most protects against this prospect. None of this discussion, of course, denies either the possibility or the desirability of crafting carefully circumscribedexceptions to First Amendment norms of viewpoint neutrality, and in the last section of this Essay, I briefly consider whether and how this task might be accomplished. Perhaps more important, none of this discussion gainsays the possibility of responding to the harms of pornography and hate speech through measures that do not contravene these norms. It is surely these measures, viewed from a pragmatic perspective, that stand the best chance of succeeding. And it usually will be these measures that pose the least danger to free speech principles. I turn, then, to a consideration of such proposals, less with the aim of making specific recommendationsthan with the aim of injecting new questions into the debate on hate speech and pornographyregulation.
II. NEW APPROACHES
I canvass here four general approaches;each is capable of encompassing many specific proposals. The four approaches are, in order: (1) the enactment of new, or the stricter use of existing, bans on conduct; (2) the enactment of certain kinds of viewpointneutral speech restrictions; (3) the enhanced use of the constitutionally unprotected category of obscenity; and (4) the creation of carefully supported and limited exceptions to the general rule against viewpoint discrimination. The proposals I outline within these approaches are meant to be illustrative, rather than exhaustive. Many fall well within constitutional boundaries; others test (or, with respect to the fourth approach, directly challenge) the current parameters. The latter proposals raise hard questions relating to whether they (no less than the standard viewpoint-based regulation) too greatly subvert principles necessary to a system of free expression. I will touch on many of these questions, although I cannot give them the extended treatment they merit. A. Conduct The most obvious way to avoid First Amendment requirements is to regulate not speech, but conduct. Recently, some schol-
884
[60:873
ars have sought to meld these two together.31Speech is conduct, they say, because speech has consequences (speech, that is, "does" something); or conduct is speech because conduct has roots in ideas (conduct, that is, "says" something). I use these terms in a different sense. When "conduct"becomes a synonym for "speech" (or "speech" for "conduct"), the command of the First Amendment becomes incoherent;depending on whether the paradigmof conduct or speech holds sway, government can regulate either almost everything or almost nothing. The speech/conduct line is hard to draw, but it retains much meaning in theory, and even more in practice. When I say "conduct,"then, I mean acts that, in The governpurpose and function, are not primarilyexpressive.32 ment can regulate such acts without running afoul of the First Amendment.33 Here, I discuss two specific kinds of conduct regulation: the continued enactment and use of hate crimes laws and the increased application of legal sanctions for acts commonly performed in the making of pornography. The typical hate crimes law, as the Supreme Court unanimously ruled last Term, presents no First Amendment problem.34 Hate crimes laws, as usually written, provide for the enhancement of criminal penalties when a specified crime (say, assault) is committed because of the target's race, religion,or other listed status.35 These laws are best understood as targeting not speech, but acts-because they apply regardless whether the discriminatory conduct at issue expresses, or is meant to express, any sort of message. In this way, hate crimes laws function precisely as do other discriminationlaws-for example, in the sphere of employment.36
31 See MacKinnon, Feminism Unmodified at 129-30, 193-94 (cited in note 13); Charles R. Lawrence III, If He Hollers Let Him Go: Regulating Racist Speech on Campus, 1990 Duke L J 431, 438-44. 32 The approach, in focusing on expressive quality, is similar to the analysis that Cass Sunstein presents in these pages. See Words, Conduct, Caste, 60 U Chi L Rev at 807-09 (cited in note 17). See also Amar, 106 Harv L Rev at 133-39 (cited in note 27). Of course, as sketched here, the definition begs all kinds of questions about when acts, either in purpose or in function, are primarily expressive. 33 So, for example, it goes without saying that the City of St. Paul could have proceeded against the juvenile offenders in R.A. V. through the law of trespass. See R.A.V., 112 S Ct at 2541 n 1 (listing other statutes under which the offenders could have been punished). S4 Wisconsin v Mitchell, 113 S Ct 2194 (1993). 35 See, for example, Cal Penal Code ? 422.7 (West 1988 & Supp 1993); NY Penal Law ? 240.30(3) (McKinney Supp 1993); Or Rev Stat ? 166.165(1)(a)(A) (1991); Wis Stat Ann ? 939.645 (West Supp 1992). 36 The Supreme Court in Mitchell noted the precise analogy between Title VII and the hate crimes statute at issue in the case. See 113 S Ct at 2200. It is noteworthy that both
1993]
885
When an employer fires an employee because she is black, the government may impose sanctions without constitutional qualm. This is so even when the discharge is accomplished (as almost all discharges are) through some form of expression, for whatever expression is involved is incidental both to the act accomplished and to the government's decision to prevent it.37 The analysis ought not change when a person assaults another because she is black, once again even if the conduct (assault on the basis of race) is accompanied by expression. A penalty enhancement constitutionally may follow because it is pegged to an act-a racially-basedform of disadvantage-that the state wishes to prevent, and has an interest in preventing, irrespective of any expressive component. In other words, in the assault case, no less than in the discharge case, the government decides to treat race-based acts differently from similar non-race-based acts; and in the assault case, no less than in the discharge case, this decision-a decision to prevent disproportionate harms from falling on members of a racial group-bears no relation to whether the race-based act communicates a message. Thus might end the constitutional analysis. Perhaps, however, this argument is not quite so easy as I have made it out to be. It might be said, in response, that racially-based assaults, more often than racially-based discharges, are committed in order to make a statement. If this is true, a penalty enhancement not only will restrict more speech incidentally, but also may raise a concern that the government is acting for this very purpose. Or perhaps it might be said, more generally, that the use of a discriminatory motive to define an act, even supposing the act has no expressive component, at times may be highly relevant to First Amendment analysis: consider, for example, a penalty enhance-
laws apply not only irrespectiveof whether the discriminationat issue expresses a message, but also irrespective of whether the discriminationis caused by particularbeliefs. If, for example, discrimination laws prohibited discharges or assaults motivated by racial hatred-rather than simply based on race-they would pose a very different, and seemingly severe, First Amendment problem. 37 Cass Sunstein makes this point in Words,Conduct,Caste, 60 U Chi L Rev at 827-28; his phrasingis that in such a case, the communicationis merely evidence of, or a means of committing, an independently unlawful act. Professor Sunstein, however,appears to think that this analysis fails to cover hate crimes,because there the state's interest arises from the expressivenature of the conduct.As stated in the text, I do not believe this to be the case. A state has a legitimate interest in preventing, say, assaults on the basis of race, even when they are wholly devoid of expression. The interest is the same as the one in preventing dischargeson the basis of race;it is an interest in eradicatingracially-basedforms of disadvantage generally, whether or not accompaniedby communicationof a message.
886
[60:873
ment provision applicable to persons who obstruct voting on the basis of a voter's affiliation with the Republican Party. But both of these objections seem to falter on further consideration of the nature of hate crimes regulationand the governmental interest in it. The voting obstruction law I have hypothesized (no less than a hate crimes law) applies to conduct regardless of whether it has expressive content, but the government'sinterest in the law always in a certain sense relates to expression:it is difficult to state, let alone give credence to, any interest the government could have, other than favoring or disfavoringpoints of view, for specially penalizing voting obstruction based on affiliation with a In the case of hate crimes laws, by conparticularpolitical party.38 the not trast, government only is regulating acts irrespective of their expressive component, but also has a basis for doing so that is unrelated to suppressing (or preferring)particularviews or expression-the interest, once again, in preventing conceded harms from falling inequitably on members of a particular racial group. In such a case, the regulationshould be found to accord with First Amendment requirements,notwithstandingthat it incidentally affects some expression.As the Court in R.A.V. noted, in referringto employment discriminationlaws, "Wherethe governmentdoes not target conduct on the basis of its expressive content,"-and where, we might add, the government,in regulatingconduct, has a credible interest that is unrelated to favoring or disfavoring certain ideas or expression-"acts are not shielded from regulationmerely because they express a discriminatoryidea or philosophy."S3 In accord with this reasoning,communitiesshould be able not only to impose enhanced criminal sanctions on the perpetratorsof hate crimes, but also to provide special tort-based or other civil remedies for their victims. One of the accomplishmentsof the antipornographymovement has been to highlight the benefits of using the civil, as well as the criminal,laws to deter and punish undesir-
38 The hypotheticalvoting law might seem very differentif enhancedpenaltiesapplied to obstructionbased on the voter's affiliationwith any political party, ratherthan with the RepublicanParty alone. In enactingthis broaderlaw, the state could have determinedthat it had an interest in protectingpersonsfrom sufferingdisproportionate harmas a result of their political views, analogousto the interest in protectingpersonsfrom sufferingdisproportionateharm as a result of their race. Under the analysissuggestedin the text, this new voting law would meet constitutionalstandards because it applies regardlesswhether the conductcommunicates a messageand becausethe government now has a credibleinterest in the law not related to favoringor disfavoringparticularviewpointsand messages. 39 112 S Ct at 2546-47.
1993]
887
able activity.40 Civil actions involve fewer procedural safeguards for the defendant, including a much reduced standard of proof; as important, they may give greater control to the victim of the unlawful conduct than a criminal prosecution ever can do. Communities therefore should consider not merely the enactment of hate crimes laws, but also the provision of some kind of "hate torts" remedies. And in determining the scope of all such laws, communities should consider the manner in which the laws apply to crimes or civil violations committed on the basis of sex, which now often fall outside the compass of hate crimes statutes. To address the harms arising from pornography,the government has numerous available mechanisms that regulate not speech, but conduct. At an absolute minimum, states can prosecute actively, under generally applicable criminal laws, the sexual assaults and other violent acts so frequently committed against women in the making of pornography. Similarly, as Judge Easterbrook suggested in Hudnut, states may specifically make illegal (if they have not already) the use of fraud, trickery, or force to induce people to perform in any films, without regard to viewpoint.41Extensive regulation of such practices is the lot of many industries; the visual media surely are not entitled to any special exemption. With respect to regulatory effects of this kind too, responses based on the criminal law can be supplemented by enhanced tort remedies.42 A much more questionable means of deterring the production of pornographicworks would be to press into service laws regulating prostitution, pimping, or pandering. In one recent case, an Arizona court upheld, against First Amendment challenge, the use of prostitution and pandering statutes against a woman who managed and performed in a sex show.43 The court reasoned, consistent with established First Amendment doctrine, that the prosecutions were
40 See CatharineA. MacKinnon,Pornography,Civil Rights, and Speech, 20 Harv CRCL L Rev 1, 29 n 52 (1985). 41 771 F2d at 332. 42 For a discussion of whether the government,in addition to banning the conduct itself, may prohibit the disseminationof speech producedby means of this unlawfulconduct, see text accompanyingnotes 55-61. 43 Arizona v Taylor, 167 Ariz 429, 808 P2d 314, 315-16 (1990). The state's prostitution statute prohibited "engagingin or agreeing or offering to engage in sexual conduct with another person under a fee arrangementwith that person or any other person."Id. The use of statutes of this kind against women who merely performin pornographyraises a special concern: such prosecutions make a criminal of the very victim of exploitative practices. Moreover,these prosecutionsmay have little value: they are likely to deter the production of pornographyfar less well than prosecutingthe actual pornographer under pimping, pandering,or other similarstatutes, which essentially prohibitthe hiringof personsto engage in sexual practices.
888
[60:873
permissible because even if the show had expressive content, the state had acted under statutes directed at conduct in order to further interests unrelated to the suppression of expression." The same argument could be made whenever the government acts against a pornographerunder a sufficientlybroad pimping or pandering statute, so long as the prosecutionwere based on a significant interest unrelated to speech, such as the prevention of sexual exploitation. The problem with this analysis lies in its potential scope: many films that no one would deem pornographiccontain sexual conduct by hired actors and thus fall within the very same statutes. Notwithstanding all I have said above, even the neutral application of a law that is not itself about speech might in some circumstancesviolate the First Amendment. (Consider,to use an extreme example, an environmentallaw imposing a ban on cutting down trees, as applied to producersof books and newspapers.)In all probability, the use of pimping and pandering statutes in the way I have just considered suffers from this constitutional defect, given the potential for applying such statutes to large amounts of speech at the core of constitutional protection. Those favoring the direct regulation of pornography often charge that relying exclusively on bans on conduct-most notably, a ban on coerced performances-would allow abuses currently Such committed in the manufactureof pornographyto continue.45 if will even have less approaches, determinedly enforced, certainly effect than banning pornographyaltogether. But once again, the most sweeping strategies also will be the ones most subject to constitutional challenge and the ones most subversive of free speech principles. An increased emphasis on conduct, rather than speech, provides a realistic, principled, and perhaps surprisinglyeffective alternative. B. Viewpoint-NeutralRestrictions The Supreme Court often has said that any speech restriction based on content, even if not based on viewpoint, presumptively violates the First Amendment.46 But rhetoric in this instance is
44 Id at 317. The key case supporting this analysis is United States v O'Brien, 391 US 367 (1968), in which the Court approved the use of a statute prohibiting any knowing destruction of a Registration Certificate, purportedly enacted to further the efficient operation of the draft, against a person who had burned his draft card as part of a political protest. 45 See, for example, Cass R. Sunstein, Neutrality in Constitutional Law (With Special Reference to Pornography, Abortion, and Surrogacy), 92 Colum L Rev 1, 23-24 (1992). 46 See, for example, Police Department of Chicago v Mosley, 408 US 92, 95-96 (1972); Simon & Schuster, Inc. v Members of the New York State Crime Victims Board, 112 S Ct
1993]
889
semi-detached from reality. The Court, for example, sometimes has upheld regulations based on the subject matter of speech.47And the Court in several cases has approved restrictions on non-obscene but sexually explicit or scatological speech.48Cases of this kind raise the possibility of eradicating the worst of hate speech and pornographythrough statutes that, although based on content, on their face (and, to the extent possible, as applied) have no viewpoint bias. One potential course is to enact legislation, or use existing legislation, prohibiting carefully defined kinds of harassment, threats, or intimidation, including but not limited to those based on race and sex. For example, in considering the St. Paul ordinance, the Court in R.A.V. noted that the city could have achieved "precisely the same beneficial effect" through "[a]n ordinance not limited to the favored topics"49-that is, through an ordinance prohibiting all fighting words, regardless whether based on race, sex, or other specified category. An ordinance of this kind would have presented no constitutional issue at all given the Court's prior holdings that fighting words are a form of unprotected expression.50 A law prohibiting, in viewpoint-neutral terms, not merely fighting words but other kinds of harassment and intimidation would (and should) face greater constitutional difficulties, relating most notably to overbreadth and vagueness; but a carefully drafted statute might well surmount these hurdles, and such a law surely would not be subject to the selectivity analysis of R.A.V. Viewpoint-neutral laws of this kind-whether framed in terms of fighting words
501, 508-09 (1991); ConsolidatedEdison Co. of New Yorkv Public Service Commissionof New York,447 US 530, 536 (1980). 47 See, for example, Burson v Freeman, 112 S Ct 1846 (1992); Greer v Spock, 424 US 828 (1976); CBS v Democratic National Committee,412 US 94 (1973). See generallyGeoffrey R. Stone, Restrictions of Speech Because of its Content: The Peculiar Case of SubjectMatter Restrictions, 46 U Chi L Rev 81 (1978). R.A.V. might be thought to treat subject matter restrictions with the same distrust shown to viewpoint restrictions:the technical holding of the Court was that the St. Paul ordinancefacially violated the Constitution"in that it prohibits otherwise permitted speech solely on the basis of the subjects the speech addresses." 112 S Ct at 2542. But elsewhere in the opinion, the Court made clear that its true concern related to viewpoint bias. What most bothered the Court was that the subject matter restriction operated in practice to restrict speech of only particular (racist, sexist, etc.) views. See, for example, id at 2547-49. 48 See FCC v Pacifica Foundation, 438 US 726 (1978) (indecent radio broadcast); Young v American Mini-Theatres, 427 US 50 (1976) ("adult"theaters); City of Renton v Playtime Theatres, Inc., 475 US 41 (1986) (same). 49 112 S Ct at 2550. 50 See Chaplinsky v New Hampshire, 315 US 568, 572 (1942). Of course, the application of the ordinance to any particular expression might well raise serious constitutional issues relating to the permissible scope of the fighting words category.
890
[60:873
or in some other manner-might be especially appropriatein communities (such as, perhaps, educational institutions) whose very purposes require the maintenance of a modicum of decency.6" Another approach,relevant particularlyto pornography,could focus on regulating materials defined in terms of sexual violence. At first glance, R.A.V. and (especially) Hudnut seem to doom such efforts, but this initial appearancemay be deceptive. The problem in Hudnut involved the way the ordinance under review distinguished between materials presenting women as sexual equals and materials presenting women as sexual subordinates: two works, both equally graphic, would receive different treatment because of different viewpoints.62This problem, the court suggested, would not arise if a statute instead were to classify materials accordingto their sexual explicitness.63 Indeed, the Supreme Court already has said as much by treating as non-viewpoint-based(and sometimes upholding) regulations directed at even non-obscene sexually graphic materials.64If a regulation applying to sexually explicit materials does not raise concerns of viewpoint bias, perhaps neither does a regulation applying to works that are both sexually explicit and sexually violent. One counterargumentmight run that the reference to sexual violence in this hypothetical statute would function simply as a code word for a disfavored viewpoint: sexually violent materials present women as subordinates;sexually non-violent materialspresent women as equals; hence, the law replicates in covert language the faults of the MacKinnon-Dworkin ordinance.But this response strikes me as flawed, because many non-violent works present women as sexual subordinates,and some violent materialsmay not (violence is not necessarily a synonym for non-equality).The question is by no means free from doubt-much depends on how far the Court will or should go to find viewpoint discriminationin a facially neutral statute-but framing a statute along these lines seems worth consideration.
Post, Racist Speech, Democracy,and the First Amendment,32 Wm & MaryL Rev 267, 317-25 (1991), for a generaldiscussionof the compatibilityof speech regulation with the objectivesof higher education.
52
"1 See RobertC.
63 Id at 332-33. b4 See note 48 and accompanying text. The Courthas failed to indicatepreciselywhen will meet constituregulationsof this kind, even assumingthey are not viewpoint-based, tional standards.All of the regulationsupheld by the Court have involved not complete bans, but more limited restrictions.A law foreclosingsuch speech entirelywould raise constitutional concernsof greaterdimension.
See 771 F2d at 328.
1993]
891
Finally, and once again of particularrelevance to pornography, the Constitution may well permit direct regulation of speech, if phrased in a viewpoint-neutral manner, when the regulation responds to a non-speech related interest in controlling conduct involved in the materials' manufacture. Assume here, as discussed above, that the government has a strong interest in regulating the violence and coercion that often occurs in the making of pornography.55Does it then follow that the government may punish the distribution of materials made in this way as well as the underlying unlawful conduct? The Supreme Court's decision in New York v Ferber56suggests an affirmative answer. In Ferber, the Court sustained a statute prohibiting the distribution of any material depicting a sexual performance by a child, primarily on the ground that the law arose from the government'sinterest in preventing the conduct (sexual exploitation of children) necessarily involved in making the expression. Similarly, it would appear, the government may prohibit directly the dissemination of any materials whose manufacture involved coercion of, or violence against, participants. The Hudnut Court specifically indicated that such a statute would meet constitutional requirements.57 Important questions remain unanswered with respect to this approach, for there are almost surely limits on the principle that the government may engage in viewpoint-neutral regulation of speech whenever it has an interest in deterring conduct involved in producing the expression. The principle itself, in addition to explaining Ferber, may explain such disparate outcomes as the ability of a court to enjoin the publication of stolen trade secrets and to award damages for the unapproved publication of copyrighted material.58 But some hypothetical applications of the principle sugthe for a boundary line. For example, could the governneed gest ment prohibit all speech whose manufacture involved violations of the Fair Labor Standards Act? Surely such a statute would violate the Constitution. Or, to use another sort of case, could the government prohibit the distribution of all national security information stolen from government agencies? An affirmativeanswer would require overruling the Pentagon Papers case.59The question arises,
66
See text accompanying notes 41-42. 56 458 US 747 (1982).
See 771 F2d at 332-33. See Harper & Row, Publishers, Inc. v Nation Enterprises, 471 US 539 (1985). 69 See New YorkTimes Co. v United States, 403 US 713 (1971). I thank Geof Stone for suggesting this example.
67 68
892
[60:873
then, how to separate permissible from impermissibleapplications of the principle. I am not sure that any factor, or even set of factors, can serve to explain fully all the cases mentioned. Some relevant considerations, however, might include the value of the speech at issue, the magnitude of the harm involved in producing the speech, the extent to which prohibitingthe speech is necessary to prevent the harm from occurring,and the extent to which the With reexpression itself reinforcesor deepens the initial injury.6? spect to all of these considerations, the prohibition of materials whose manufactureinvolves sexual violence seems similar enough to the ban in Ferber to suggest that the regulation,while deterring the worst forms of pornography,still would satisfy First Amendment standards.6' C. Obscenity The government can also regulate sexually graphic materials harmful to women by using the long-established category of obscenity. This approach to regulating such materials has come to femiassume the aspect of heresy in the ranks of anti-pornography nism. Those who have argued for regulating pornographyhave stressed the differences, in rationale and coverage, between bans
60 The Ferber Courtviewed the harm involvedin manufacturing child pornography as great and the value of the resultingexpressionas usually,thoughnot always,slight. See 458 US at 757-58, 762-63, 773-74. With respect to the necessity of prohibitingnot merely the unlawfulconduct, but also the speech itself, the Ferber Courtstated that "the distribution must be closed if the productionof materialwhich requires networkfor child pornography the sexual exploitation of children is to be effectively controlled."Id at 759. Finally, the Ferber Court noted that "the materialsproducedare a permanentrecordof the children's Id. participationand the harm to the child is exacerbatedby their circulation."
61
The Supreme Court's decision in City of Renton v Playtime Theatres, 475 US 41
(1986), might be taken to suggest-although, I believe, wrongly-a furtherextensionof the that the government argument: may prohibitthe distributionof materialseven substantially correlatedto unlawfulconductin manufacture, so long as the definitionof these materialsis In Renton, the Courtupheld the regulationof adult motionpicturetheaviewpoint-neutral. ters on the ground that such theaters generallycorrelatewith a rise in crime in the surId at 50. The Courtdeclinedto requirea showingthat any particuroundingneighborhood. lar movie theater in fact producedthese results. Similarly,a statute regulatinga categoryof speech that is highly correlated with coercion of, or violence against, women might be thought to pass constitutionalmuster even if a particularinstance of that speech did not involvecoercionor violence.This line of argument,however,takes what I believeitself to be a problematicdecisionmuch too far. Crucialto the Renton holdingwas the limited scope of the regulationunder review:it zoned adult theaters, but did not prohibitthem. Id at 53. A total ban on speech, based on a mere correlationbetweenthe speech and unlawfulconduct (even if the conduct, as in Renton and here, stemmed from something other than the speech's communicative effects), would raise constitutional concerns of much greater magnitude.
1993]
893
on the pornographic and bans on the obscene. It is said that obscenity law focuses on morality, while pornography regulation focuses on power.62 It is said that offensiveness and prurience (two of the requirements for finding a work obscene) bear no relation to sexual exploitation.63It is said that taking a work "as a whole," as obscenity law requires, and exempting works of "serious value," as obscenity law does, ill-comports with the goal of preventing harm to women.64I do not think any of this is flatly wrong, but I do wonder whether these asserted points of difference-today, even if not in the past-suggest either the necessity or the desirability of spurning the obscenity category. My doubts began in the midst of first teaching a course on free expression. In keeping with the prevailing view, I rigidly segregated the topics of obscenity and pornography. (If I recall correctly, I taught commercial speech in between the two.) In discussing each, I iterated and reiterated the distinctions between them, in much the terms I have just described. I think I made the points clearly enough, but my students resisted; indeed, they could hardly talk about the one topic separately from the other. In discussing obscenity, they returned repeatedly to the exploitation of women; in discussing pornography,of course, they dwelt on the same. Those who favored regulation of pornography also favored regulation of obscenity-at least as a second-best alternative. Those who disapproved regulation of pornography also disapproved regulation of obscenity. Perhaps it was a dense class or I a bad teacher, but I think not; rather, I think the class understood-or, at the very least, unwittingly revealed-something important. Even when initially formulated, the current standard for identifying obscenity was justified in part by reference to real-world harms. To be sure, the Supreme Court, in its fullest statement of the rationale for establishing the category of obscenity, spoke of the need "to protect 'the social interest in . . . morality'" and, what is perhaps the same thing, of the need "'to maintain a decent society . . . ' "65 Here, the Court appeared to stress a version of morality divorced from tangible social consequences and related to simple sentiments of offense or disgust. But the Court also spoke
See MacKinnon,Feminism Unmodified at 147 (cited in note 13). See id at 174-75;Sunstein, 92 Colum L Rev at 20-21 (cited in note 45). 64 See MacKinnon,Feminism Unmodified at 174-75 (cited in note 13). 66 Paris Adult Theatre I v Slaton, 413 US 49, 59-60, 61 (1973) (emphasis deleted), quoting Jacobellis v Ohio, 378 US 184, 199 (1964) (Warrendissenting), and Roth v United States, 354 US 476, 485 (1957).
62
63
894
[60:873
of-indeed, emphasizedjust as strongly-the "correlationbetween obscene material and crime" and, in particular,the correlationbetween obscene materials and "sex crimes."66 This concern too may reflect a notion of morality, but if so, it is a morality rooted in material harms.67And although some of the specific harms then perceived might now appear dated-the Court was thinking as much of unlawful acts involving "deviance"as of unlawful acts involving violence-still the Court understoodthe obscenity category as emerging not merely from a body of free-floating values, but from a set of tangible harms, perhaps including sexual violence.68 Much more important is the way conceptions of obscenity have evolved since then, in part because of the anti-pornography movement itself, in part because of the deeper changes that movement reflects in public attitudes and morals. This shift in understanding, I think, accounted for my classroom experience. It is hard to test a proposition of this sort, but I will hazard it anyway: one of the great (if paradoxical)achievements of the anti-pornography movement has been to alter views on obscenity-to transform obscenity into a category of speech understood as intimately related, in part if not in whole, to harms against women.69 Surely, such a change in perception should come as no great surprise. It would be the more astonishing by far if obscenity were viewed today as obscenity was viewed two decades ago, when the current constitutional standard was first announced. A doctrinal test does not so easily freeze public understandings,especially when the test in part relies (as the obscenity test does) on community standards and consciousness.70 Views of obscenity, in other words, are not
O. Conkle,Harm, Morality, and Feminist Religion:Canada'sNew-But Not So New-Approach to Obscenity,10 Const Comm105, 123-24(1993),for discussionof these two kinds of morality (offense-based and harm-based) as reflected in obscenity doctrine. 68 For this reason, I think CatharineMacKinnon's statement that obscenity is "ideational and abstract,"rather than "concreteand substantive,"representssomethingof an even as appliedto the initial understanding and formulation of the category. overstatement, See MacKinnon,Feminism Unmodifiedat 175 (cited in note 13). 69 One interestingproof (and product)of this reconceptualization is SenatorMitch McConnell'sproposedlegislationgrantingthe victim of a sexual offense a right to claim damages from the distributorof any obscene work deemed to have contributedto the crime. Victims'Compensation Act of 1991,S 1521, 102dCong,1st Sess (Jul 22, 1991). Pornography Whateverthe merits of this legislation,which raises seriousconcernson numerousgrounds, it clearly presupposesa link between obscenity and sexual violence. 70 The obscenity standard asks whether the average person, applying contemporary communitystandards,would find a work prurientand offensive in its depiction of sexual
66 413 US at 58-59. 67 See Daniel
1993]
895
static, and they may have evolved in such a way as to link obscenity with harms to women. Now it might be argued, in response to this claim, that so long as the formal test for determining obscenity remains the same, this reconceptualization of obscenity will avail women little, because the test's focus on prurience and offensiveness will prevent new understandings from affecting judicial outcomes. But this response seems to ignore the subtle and gradual ways law often develops. As prosecutors, juries, and judges increasingly adopt this new view of obscenity, enforcement practices and judicial verdicts naturally will come to resemble, although not to replicate, those that would obtain under an anti-pornographystatute. There is in fact a substantial overlap between the categories of obscenity and pornography: most of the worst of pornography(materials with explicit and brutal sexual violence) meets the obscenity standard. As public perceptions continue to change, the application of the obscenity standard increasingly will focus on the materials causing greatest harm to women; nor need this development reflect any illegitimate acts of prosecutorial discretion.71 Moreover,this new focus may over time reshape, in a desirable manner, even the governing legal standard for determining obscenity. Doctrinal adjustments and reformulationsof existing low-value categories of speech may well-and should-occur more readily than the creation of whole new categories, especially when the proposed new categories incorporate clear viewpoint bias. So, for example, the current obscenity test's requirement that materials be patently offensive may disintegrate in light of new understandings about the harms the obscenity category principally should address. This evolution of obscenity law recently has occurred in Canada, where the Supreme Court, responding to increased evidence and altered perceptions of harm to women, made sexual violence rather than sexual offensiveness the keystone of the obscenity category.7 Efforts to redefine the obscenity category in this manner-a redeficonduct. It also asks whether the work lacks serious literary, artistic, political, or scientific value. See Miller v California, 413 US 15, 24 (1973). 71 If prosecutorsdetermine to enforce obscenity laws only against materialswith a certain viewpoint, the resulting actions would be no less problematic than the MacKinnonDworkin statute itself. But this result is hardly the only one that could be produced by changingpublic norms. For example, as noted earlier and discussed again below, a focus on sexual violence arguablyis not viewpoint-biased.See text accompanyingnotes 52-54 and 74. Thus, to the extent that prosecutorsenforce obscenity laws strictly against sexually violent materials that fall within the obscenity category, their acts would not violate the R.A.V. proscriptionof preferringsome viewpoints to others within a low-value category. 72 See Regina v Butler and McCord, [1992] 1 SCR 452, 134 NR 81, 108-18 (Canada).
896
[60:873
nition that, consistent with much First Amendment theory, would tend to divorce speech restrictions from simple feelings of offense-should proceed in the United States as well.73 One measure along these lines that states or localities might attempt involves the special regulation of subcategoriesof obscenity that contain sexual violence. R.A.V. might seem to bar such an approach;it held, after all, that even within low-valuecategoriesof speech, such as obscenity or fighting words, the government may not make distinctions that pose a danger of viewpoint bias. I have argued above that a statute framed in terms of sexual violence may no more implicate this principle than the several statutes upheld But even if by the Court framed in terms of sexual explicitness.74 courts reject this argument,another possibility presents itself. The Court in R.A.V. stated as an exception to its broad rule that a subcategory of unprotected speech can be specially regulated if it presents, in especially acute form, the concerns justifying the exclusion of the whole category from First Amendment protection.7" It is hard to know what this exception means, especially in light of the Court's refusal to apply it to the category of race-based fighting words, which appears to pose in especially acute form the dangers giving rise to the entire fighting words category. It is no less difficult to determine what the exception should mean, given the ability to characterize in many different (and even conflicting) ways the concerns underlying any low-value category and the ease of restating those concerns with respect to any given subcategory. But given the Court'sacknowledgmentof the relationshipbetween sexual crimes and obscenity, some considerationshould be given to whether a statute focusing on the particular kinds of obscenity that most contribute to sexual violence would or should fall within the R.A.V. exception.76
of the obscenitycategorywould renderit 79 It might be arguedthat such a redefinition viewpoint-basedand therefore inconsistentwith the First Amendment.This argumentdepends first on the propositionthat a statute framedin terms of sexual violenceis viewpointnotes 52-54.As important,the argubased,which I have discussedin the text accompanying ment depends on the proposition that the obscenity category is not now viewpointbased-in other words,that it does not now constitute some kind of exceptionto the rule of viewpointneutrality.This propositionis difficultto maintaingiven the obscenitytest's reliance on community standards of offensiveness.See Sunstein, 92 Colum L Rev at 28-29 (cited in note 45). As between an obscenity doctrine that focuses on sexual prurienceand offensivenessand an obscenity doctrine that focuses on sexual prurienceand violence, the formerwould appearto pose the greaterdangerof viewpointbias. 74 See text accompanying notes 52-54 and notes 71 and 73. 76 112 S Ct at 2545-46. 76 The Courtwrote,for example,that "a State may choose to regulateprice advertising in one industry but not in others, because the risk of fraud (one of the characteristics of
1993]
897
The key point here is that regulation of obscenity may accomplish some, although not all, of the goals of the anti-pornography movement; and partly because of the long-established nature of the category, such regulation may give rise to fewer concerns of compromising First Amendment principles. Even for those who think that the obscenity doctrine is in some sense a second-best alternative, it represents the first-best hope of achieving certain objectives. And the obscenity doctrine itself may benefit by transformative efforts, as these efforts bring the doctrine into greater accord with the harm-based morality of today, rather than of twenty years ago. D. Exceptions to Viewpoint Neutrality The final approach I will discuss, although far more briefly than it deserves, involves crafting arguments to support explicit exceptions to the rule against viewpoint discrimination for pornography or hate speech. As noted earlier, exceptions to this rule do exist, but without any clear rationale; the Court, in upholding viewpoint discriminatoryactions, simply has ignored their discriminatory nature. We know, from the decision in R.A.V. and the affirmance of Hudnut, that the Court will follow no such course of studied inattention with respect to pornographyor hate speech: in both cases, the presence of viewpoint discrimination was considered-and was declared dispositive. The question, then, arises: Is it possible to make a convincing argument to the contrary? Is it possible, that is, to accept viewpoint neutrality as a general principle, but to support an exception to that principle either for pornography or for hate speech? The challenge here is to explain in credible fashion what makes one or two or three viewpoints (or one or two or three instances of viewpoint discrimination) different from all others-sufficiently different to support an exception and sufficiently different to ensure that the exception retains "exceptional" status. I cannot here provide the answer to that question. Instead, I will confine myself to some general observations about what considerations might be relevant to the inquiry. Two factors necessary (but, I will argue, generally insufficient) for departing from the norm of viewpoint neutrality are (1) the
commercialspeech that justifies deprivingit of full First Amendmentprotection) . . . is in its view greaterthere." Id at 2546. So too, it might be said, a State may choose to regulatein a special mannersexually violent obscenity because it poses a greaterrisk of contributingto sexual crimes-one of the characteristicsof obscenity that justifies deprivingit of full First Amendment protection.
898
[60:873
seriousness of the harm the speech causes, and (2) the "fit" between the harm and the viewpoint discriminatorymechanismchosen to address it. The first considerationhas an obvious basis: to the extent a viewpoint causes insignificant harm, the state's decision to suppress that viewpoint must rest not on legitimate reasons but on mere dislike of the idea at issue. The second consideration is related and not much more mysterious: when the government restricts a viewpoint, but the viewpoint is not coextensive with the harm allegedly justifying the governmentalaction, we may wonder (once again) whether the action is in fact motivated by simple distaste for the message. I have no doubt that a regulationof pornography and hate speech would satisfy the first inquiry, and little doubt that such a regulationcould be carefullyenough constructed to satisfy the second. Is that, however, sufficient? I think not. Assume, for example, a carefully crafted regulation of abortion advocacy, counseling, or referral (the category of speech involved in Rust v Sullivan77), designed to reduce the incidence of abortions. Proponents of the regulation might urge that the law is precisely crafted to reduce the significant harms stemming from abortion; hence the law satisfies the two inquiries set forth above. I presume this outcome would strike many as irretrievably wrong.But, some opponents of the regulationmight contend, the example fails to prove my larger point because the "harms"in the hypothetical case (howeverserious some might find them) are in fact widely contested and for that reason cannot form the basis of viewpoint regulation.These opponents might contrast a precisely crafted regulation of pro-smokingspeech, designed to reduce the frequency of tobacco use. In that case, the harms are not contested; hence the regulation can go forward.The contrast here has much intuitive appeal, and I am not at all sure it has nothing to teach us. But this general line of reasoning makes the protections of the First Amendment weakest at the very point where views are the most unorthodox and unconventional. And even if I am wrong to think this result upside-downand unacceptable, another question would follow: Are not the harms caused by pornographyand hate speech-characterized most generallyas racial and sexual subordination-also very much contested? If they were not, the debate over hate speech and pornographymight not have reached so intense a level.
77
111 S Ct 1759, 1765 (1991).
1993]
899
Assuming, then, that harm and "fit" cannot alone justify viewpoint discrimination, perhaps the addition of low-value speech can do so. In other words, if legislators can make the case that speech leads to harm, if the speech regulated correlates precisely with that harm, and if the speech is itself low-value, then any viewpoint discrimination involved in the regulation becomes irrelevant.78 At first of R.A.V. this course, glance, definitively rejected argument: the very holding of that case was that even within a low-value category of speech, viewpoint discrimination is generally prohibited. So, to use one of the Court's hypotheticals, the government may proscribe libel, but may not proscribe only libel attacking the government; or, to use something near the actual case, the government may prohibit fighting words, but may not prohibit only racist fightBut what, then, are we to make of a category like obing words.79 scenity-an entire low-value category (rather than a subdivision thereof) that seems to incorporate some viewpoint bias?80Could it possibly be the case that viewpoint discrimination built into the very definition of a low-value category is permissible, whereas viewpoint discrimination carving up a neutrally defined low-value category is not? The proposition is perhaps less silly than it appears, for the latter, but not the former, lacks the precise "fit" that I above termed necessary for viewpoint regulation. When the Court establishes a low-value category, such as obscenity, it determines that the harms caused by the covered speech so outweigh its (minuscule) value that regulation of the speech, even if viewpoint discriminatory, will be permitted. The Court, in effect, predecides that regulation of the entire category will arise not from governmental hostility to the ideas restricted, but rather from a neutral decision based on harms and value; the viewpoint bias will occur as a mere byproduct of the fact that only the restricted ideas cause great harms and have sparse value. This predetermination insulates the government from a charge of viewpoint bias when the government regulates the entire category. But the establishment of a low-value category has no such effect when the government regulates within the category on the basis of a viewpoint extraneous to the cate78 I take Cass Sunstein to be making something like this argumentin these pages. See 60 U Chi L Rev at 829 (cited in note 17). 79 112 S Ct at 2543 & n 4. The actual ordinance, as construed, prohibited race-based fighting words (discriminatingby subject matter), but the Court arguedthat this restriction operated in practice in the same way as an ordinancebanningracist fightingwords (discriminating by viewpoint). See id at 2547-48. 80
See notes 13 and 73.
900
[60:873
gory's boundaries.In that case, there is reason to suspect that the government is acting not for the reasons already found by the Court to be legitimate, but rather out of hostility to a message. The critical failure in such a regulationrelates to "fit":because the regulation is underinclusive-because it does not regulate all speech previously determined to cause great harm and have no value-the concern arises that the governmenthas an illegitimate motive. Hence, to say, as the Court did in R.A.V., that the government may not engage in unrelated viewpoint discriminationwithin a low-value category-may not, for example, ban only obscenity produced by Democrats-is not to say that viewpoint may not enter into the very definition of a low-value category. Once again, in the latter case viewpoint serves as a placeholderfor a balance of harms and values found legitimate by the Court; in the former case, viewpoint serves as a warning signal that the government is acting for other reasons. But even if this distinction holds, the hard question remains: should the Court accept pornographyor hate speech as a low-value category of expression?The currentlyrecognizedcategoriesof lowvalue speech seem to share the trait, as Cass Sunstein writes, that they are neither "intended [nor] received as a contributionto social deliberationabout some issue."81 That definition offers several lessons for any regulation,concededlybased on viewpoint,either of hate speech or of pornography.In the case of hate speech, such an ordinance should be limited to racist epithets and other harassment: speech that may not count as "speech"because it does not contribute to deliberation and discussion. In the case of pornography, any ordinanceshould be limited to materialsthat operate primarily (as obscene materials operate primarily) as masturbatory devices; in addition, an explicit exception, like that in the obscenity standard, for works of serious value ought to be incorporated. Only if pornographyand hate speech are defined in this narrow mannermight (or should) the Court accept them as low-valuecategories-a classification that, it must be remembered,depends at least as much on the non-expressivequality of the speech as on the degree of harm the speech causes. In addition to all this, perhaps one other factor-the modesty, or limited nature, of the viewpoint restriction-should be considered priorto recognizinga low-valuecategoryof speech incorporating viewpoint bias. This inquiry would focus on whether the regu-
81
Sunstein, 60 U Chi L Rev at 807.
1993]
901
lation of the category wholly excises the viewpoint from the realm of public discourse or cuts off only a limited means of expressing Even the MacKinnon-Dworkinversion of anti-porthe viewpoint.82 nography legislation would do only the latter: it would prohibit not all messages of sexual subordination, but only those messages expressed in a sexually graphic manner. This feature seems critical to the establishment of any exception to the viewpoint neutrality principle. The broader the restriction, the more it will skew public discourse toward some views and away from others. And the larger the skewing effect, the greater the chances of improper governmental motivation; a wholesale, more than a marginal, restraint suggests a government acting not for neutral reasons, but out of simple hostility to the idea restricted. Of course, the inquiry into the scope of a viewpoint restriction does not lend itself to scientific precision. The matter is always one of degree, involving the drawing of a line someplace on a spectrum. The inquiry, too, is complicated by the issue whether the particular means restricted (even if technically modest) constitute the most effective way of delivering the message, such that the restriction ought to be treated as sweeping. But the haziness of the endeavor does not gainsay the need to engage in it. For a viewpoint restriction that results in excising ideas from public discourse ordinarily ought not to be countenanced-even when the restriction applies only to lowvalue speech and even when the restriction closely responds to serious harms.
CONCLUSION
The presumption against viewpoint discrimination, relied upon in Hudnut and further strengthened in R.A.V., has come to serve as the very keystone of First Amendment jurisprudence.This presumption, in my view, has real worth, in protecting against improperly motivated governmental action and against distorting effects on public discourse. And even if I assign it too great a value, the principle still will have to be taken into account by those who
I do not at all advocate here that courts considerthe modesty of a viewpointrestriction in all cases involving viewpoint regulation.Rather, I mean that courts should ask this question when the other criteria,discussed above, for departingfrom the viewpointneutrality rule have been met. This approachis similar to the one used in City of Renton v Playtime Theatres, Inc., 475 US 41, 53 (1986), in which the Supreme Court looked to the scope of the speech restrictionat issue-an inquirythe Court normallyeschews-in a case involving low-value speech. For a detailed discussion generallydisapprovingany inquiry into the modesty of a viewpoint restriction, although not consideringthe precise issue raised here, see Stone, Content Regulation and the First Amendment at 200-33 (cited in note 26).
82
902
favor any regulation either of hate speech or of pornography.I have suggested in this Essay that the regulatory efforts that will achieve the most, given settled law, will be the efforts that may appear, at first glance, to promise the least. They will be directed at conduct, rather than speech. They will be efforts using viewpoint-neutral classifications.They will be efforts taking advantage of the long-establishedunprotected category of obscenity. Such efforts will not eradicate all pornographyor all hate speech from our society, but they can achieve much worth achieving. They, and other new solutions, ought to be debated and tested in a continuing and multi-faceted effort to enhance the rights of minoritiesand women, while also respecting core principles of the First Amendment.

Regulation of Hate Speech and Pornography After RAV PDF

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Regulation of Hate Speech and Pornography After RAV PDF

Загружено:

Авторское право:

Доступные форматы

Regulation of Hate Speech and Pornography after R.A.V.

Regulation of Hate Speech and Pornography After R.A.V.

The University of Chicago Law Review

The University of ChicagoLaw Review

Regulation After R.A.V.

The University of Chicago Law Review

Regulation After R.A.V.

The University of ChicagoLaw Review

Regulation After R.A.V.

The University of ChicagoLaw Review

See Akhil Reed Amar, The Supreme Court, 1991 Term-Comment:

The Case of the

Regulation After R.A.V.

The University of ChicagoLaw Review

Regulation After R.A.V.

The University of ChicagoLaw Review

Regulation After R.A.V.

The University of ChicagoLaw Review

Regulation After R.A.V.

The University of Chicago Law Review

"1 See RobertC.

See 771 F2d at 328.

See text accompanying notes 41-42. 56 458 US 747 (1982).

The University of Chicago Law Review

The Supreme Court's decision in City of Renton v Playtime Theatres, 475 US 41

Regulation After R.A.V.

The University of Chicago Law Review

66 413 US at 58-59. 67 See Daniel

Regulation After R.A.V.

The University of Chicago Law Review

Regulation After R.A.V.

The University of ChicagoLaw Review

111 S Ct 1759, 1765 (1991).

Regulation After R.A.V.

See notes 13 and 73.

The University of Chicago Law Review

Sunstein, 60 U Chi L Rev at 807.

The University of ChicagoLaw Review

Вам также может понравиться